Journal of Applied Statistics最新文献

Multivariate meta-analysis is an efficient tool to analyze multivariate outcomes from independent studies, with the advantage of accounting for correlations between these outcomes. However, existing methods are sensitive to outliers in the data. In this paper, we propose new robust estimation methods for multivariate meta-analysis. In practice, within-study correlations are frequently not reported in studies, conventional robust multivariate methods using modified estimation equations may not be applicable. To address this challenge, we utilize robust functions to create new log-likelihood functions, by only using the diagonal components of the full covariance matrices. This approach bypasses the need for within-study correlations and also avoids the singularity problem of covariance matrices in the computation. Furthermore, the asymptotic distributions can automatically account for the missing correlations between multiple outcomes, enabling valid confidence intervals on functions of parameter estimates. Simulation studies and two real-data analyses are also carried out to demonstrate the advantages of our new robust estimation methods. Our primary focus is on bivariate meta-analysis, although the approaches can be applied more generally.

Generalised Random Tessellation Stratified (GRTS) is a popular spatially balanced sampling design. GRTS can draw spatially balanced probability samples in two dimensions but has not been used to sample higher-dimensional auxiliary spaces. This article considers applying dimensionality reduction techniques to multidimensional auxiliary spaces to allow GRTS to be used to sample them. The aim is to improve the precision of GRTS-based estimators of population characteristics by incorporating auxiliary information into the GRTS sample. We numerically evaluate two dimensionality reduction techniques for equal and unequal probability samples on two spatial populations. Multipurpose surveys are also considered. Results show that GRTS samples from these two-dimensional spaces can improve the precision of GRTS over spatial coordinates.

In this paper, we study the functional linear multiplicative model based on the least product relative error criterion. Under some regularization conditions, we establish the consistency and asymptotic normality of the estimator. Further, we investigate the optimal subsampling for this model with massive data. Both the consistency and asymptotic distribution of the subsampling estimator are first derived. Then, we obtain the optimal subsampling probabilities based on the A-optimality criterion. Moreover, the useful alternative subsampling probabilities without computing the inverse of the Hessian matrix are also proposed, which are easier to implement in practise. Finally, numerical studies and real data analysis are carried out to evaluate the performance of the proposed approaches.

Uncertainty quantification (UQ) is an active area of research, and an essential technique used in all fields of science and engineering. The most common methods for UQ are Monte Carlo and surrogate-modelling. The former method is dimensionality independent but has slow convergence, while the latter method has been shown to yield large computational speedups with respect to Monte Carlo. However, surrogate models suffer from the so-called curse of dimensionality, and become costly to train for high-dimensional problems, where UQ might become computationally prohibitive. In this paper we present a new technique, Lasso Monte Carlo (LMC), which combines a Lasso surrogate model with the multifidelity Monte Carlo technique, in order to perform UQ in high-dimensional settings, at a reduced computational cost. We provide mathematical guarantees for the unbiasedness of the method, and show that LMC can be more accurate than simple Monte Carlo. The theory is numerically tested with benchmarks on toy problems, as well as on a real example of UQ from the field of nuclear engineering. In all presented examples LMC is more accurate than simple Monte Carlo and other multifidelity methods. Thanks to LMC, computational costs are reduced by more than a factor of 5 with respect to simple MC, in relevant cases.

The Lorenz curve is a fundamental tool for analysing income and wealth distribution and inequality at national and regional levels. We utilise a one-way functional analysis of variance to decompose a time series of Lorenz curves and develop a method for producing one-step-ahead point and interval forecasts. The one-way functional analysis of variance is easily interpretable by decomposing an array into a functional grand effect, a functional row effect and residual functions. We evaluate and compare the forecast accuracy between the functional analysis of variance and three non-functional methods using the Italian household income and wealth data.

Diagnostic analyses in regression modeling are usually based on residuals or local influence measures and are used for detecting atypical observations. We develop a new approach for identifying such observations when the parameters of the model are estimated by maximum likelihood. The proposed approach is based on the information matrix equality, which holds when the model is correctly specified. We introduce a new definition of an atypical observation: one that disproportionately affects the degree of adequate specification of the model as measured using the sample counterparts of the matrices that comprise the information matrix equality. We consider various measures of distance between two symmetric matrices and apply them such that a zero distance indicates correct model specification. These measures quantify the degree of model adequacy and help identify atypical cases that significantly impact the model's adequacy. We also introduce a modified generalized Cook distance and a new criterion that uses the two generalized Cook's distances (modified and unmodified). Empirical applications involving Gaussian and beta regression models are presented and discussed.

It is not uncommon for surveys in the social sciences to ask sensitive questions. Asking sensitive questions indirectly enables collecting of the desirable sensitive information while at the same time protecting respondents' data privacy. The randomized response technique, which uses a randomization scheme to collect sensitive responses, is one common approach used to achieve this. In this paper, we propose a multivariate ordered probit model to jointly analyze binary and ordinal sensitive response variables. We also develop Bayesian methods to estimate the probit model and perform posterior inference. The proposed probit model is applied to a large-scale drug administration survey to understand the work practice and experience of staff in three hospitals in Hong Kong. Randomized response technique was adopted in this drug administration survey to maintain the anonymity of staff whose work practice may deviate from official hospital guidelines. Empirical results using the drug administration data illustrate that we can understand the experience and practice of staff members in giving medication through probit modeling. Knowing the staff's practice on giving medication can indicate what drug administration procedures the staff may not follow properly and what areas to focus on for the enhancing of drug administration.

Indirect exposure assessment based on average environmental concentrations in microenvironments and time spent in each environment has been considered an important way of assessing personal exposure to air pollutants. Using this indirect approach, the exposure simulator generates personal exposure values or the distribution of personal exposure for air pollutants. To match the simulator with the actual exposure measurements well, some calibration is necessary. However, unlike simulators generating personal exposure values, research evaluating the validity of the second type of simulator is rare. This study aims to develop a method for calibrating a simulator that generates an exposure distribution. To describe the relationship between the actual exposure measurements and the simulator, we introduce measurement error models (MEMs) and explain how the coefficients in the models can be used for calibrating the exposure distribution. We illustrate the proposed method using a Korea Simulation Exposure Model for fine particulate matter (KoSEM-PMII).

Traditional regression analysis primarily aims to describe the overall relationship between variables, often overlooking unexplainable aspects by design. Our focus is on these unexplained aspects, leveraging them to identify disparity groups with outlying behavior that deviate from the established model. We introduce a data-driven method for identifying such groups using group studentized residuals, which we term the mean squared of external studentized residuals. We apply this method to investigate disparities within healthcare markets, examining healthcare purchasing behavior and identifying the characteristics of disparity groups.

Estimation in GARMA models has traditionally been carried out under the frequentist approach. To date, Bayesian approaches for such estimation have been relatively limited. In the context of GARMA models for count time series, Bayesian estimation achieves satisfactory results in terms of point estimation. Model selection in this context often relies on the use of information criteria. Despite its prominence in the literature, the use of information criteria for model selection in GARMA models for count time series have been shown to present poor performance in simulations, especially in terms of their ability to correctly identify models, even under large sample sizes. In this work, we study the problem of order selection in GARMA models for count time series, adopting a Bayesian perspective considering the Reversible Jump Markov Chain Monte Carlo approach. Monte Carlo simulation studies are conducted to assess the finite sample performance of the developed ideas, including point and interval inference, sensitivity analysis, effects of burn-in and thinning, as well as the choice of related priors and hyperparameters. Two real-data applications are presented, one considering automobile production in Brazil and the other considering bus exportation in Brazil before and after the COVID-19 pandemic, showcasing the method's capabilities and further exploring its flexibility.